84b8682203
refactor(gateway): collapse defensive layers, drop dead loggers, compact call sites across lifecycle/monitoring/media modules
2026-09-02 19:21:46 -07:00
3238af61c1
refactor(gateway): tier tables for memory/disk pressure, shared ISO parser for drain markers, hand-compacted docstrings
2026-09-02 18:54:58 -07:00
5fa7066054
refactor(gateway): compact restart/drain/cache-pressure/monitor modules; extract hook-dir loader and capture-path indexer
2026-09-02 18:33:07 -07:00
98dccb4ca8
refactor(gateway): unify /proc + coercer helpers, merge timestamp regexes, collapse defensive layers in lifecycle/monitoring/media modules
2026-09-02 18:23:35 -07:00
659c6b9fff
refactor(gateway): simplify session, status, stream_consumer, kanban_watchers, relay adapter, hosted-room driver/discussion/replicas, shutdown/lifecycle, pairing, channel_directory, control_socket and small modules
2026-09-02 13:31:54 -07:00
4b659f0e33
fix(gateway): retry failed session database opens
2026-08-23 18:25:12 -07:00
81f60a0c84
fix(gateway): close readiness-probe SQLite connection deterministically
...
Sibling of the #69678/#69567 ledger leak class found while widening the
sweep: _probe_state_db used 'with sqlite3.connect(...)', whose context
manager only commits/rolls back and never closes, leaking one connection
(db fd) per health poll in the long-running gateway. Wrap the connection
in contextlib.closing so every probe closes deterministically.
2026-07-24 15:55:08 -07:00
6142203bd7
fix(gateway): ground readiness in live runtime state
2026-07-11 08:42:21 -07:00
f9728af5e2
feat(gateway): add authenticated runtime readiness checks
2026-07-11 08:42:21 -07:00
Teknium